Abstract A014: Combinations of previously reported biomarkers achieve improved sensitivity and specificity of detection of early stage pancreatic ductal adenocarcinoma

Abstract

Abstract A promising approach to the early detection of pancreatic ductal adenocarcinoma (PDAC) is surveillance among people at elevated risk for the disease using a blood test. Several groups have reported promising biomarkers, but no single biomarker delivers the rate of positively predicting cancers and of negatively predicting non-cancers needed for routine use. We hypothesized that combinations of previously reported PDAC biomarkers could contribute complementary detection across heterogeneous subgroups of patients and consequently improve biomarker performance. Blood plasma specimens from 427 individuals, grouped into a training set (n = 132) and a validation set (n = 295), were distributed to four different laboratories for analysis using their independently developed biomarker assays. We analyzed the training set to identify biomarker combinations that improved upon CA19-9 for identification of early stage PDAC relative to benign disease conditions, both for improved sensitivity while controlling for high specificity and improved specificity while controlling for high sensitivity. We then applied the biomarker combinations to the validation set to determine their performance. The combination of CA19-9 with the glycan biomarker CA199.STRA improved sensitivity from 0.44 sensitivity with 0.98 specificity for CA19-9 to 0.71 sensitivity with 0.94 specificity (p < 0.001, 1000-fold bootstrap), and CA19-9 combined with the protein biomarker LRG1 and CA199.STRA improved specificity from 0.16 specificity and 0.94 sensitivity for CA19-9 to 0.58 specificity and 0.89 sensitivity (p < 0.001, 1000-fold bootstrap). Notably, cross-laboratory biomarker combinations that did not include CA19-9 in the panel significantly improved upon CA19-9. The combination of CA199.STRA with the protein biomarker MUC5AC improved sensitivity over CA19-9 with 0.61 sensitivity and 0.94 specificity (p = 0.0002, 1000-fold bootstrap), and the combination of CA199.STRA and the protein biomarkers MUC5AC, THSP2, and ANG improved specificity over CA19-9 with 0.66 specificity and 0.88 sensitivity (p < 0.0001, 1000-fold bootstrap). By coordinating this biomarker study across independent laboratories, we have identified novel biomarker combinations that significantly increase the sensitivity and specificity of early stage PDAC detection and that demonstrate the future feasibility of surveillance for PDAC using combinations of serum assays. Citation Format: Brian Haab, Lu Qian, Ben Staal, Maneesh Jain, Johannes Fahrmann, Christine Worthington, Denise Prosser, Liudmila Velokokhatnaya, Camden Lopez, Runlong Tang, Mark Hurd, Gopalakrishnan Natarajan, Sushil Kumar, Lynette Smith, Samir Hanash, Surinder Batra, Anirban Maitra, Anna Lokshin, Ying Huang, Randall E. Brand, Liudmila Velokokhatnaya. Combinations of previously reported biomarkers achieve improved sensitivity and specificity of detection of early stage pancreatic ductal adenocarcinoma [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Pancreatic Cancer; 2023 Sep 27-30; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(2 Suppl):Abstract nr A014.